Journal of Hazardous Materials B99 (2003) 89105

Case study

Partitioning characteristics of targeted heavy metals

in IZAYDAS hazardous waste incinerator
Mithat Bakoglu, Aykan Karademir , Savas Ayberk
Department of Environmental Engineering, University of Kocaeli, Kocaeli, Turkey
Received 12 January 2002; received in revised form 24 December 2002; accepted 29 December 2002

Abstract
Partitioning of eight targeted heavy metals (Cr, Mn, Cu, Pb, Sn, Co, Ni and Zn) was carried out
during five trial burns in Izmit hazardous and clinical waste incinerator (IZAYDAS). Metal contents
of the original wastes and their concentration in the bottom ash (BA), fly ash (FA), filter cake (FC)
and flue gas were determined. Partitioning behavior of metals during the two-stage incineration was
evaluated with respect to physico-chemical properties of feed waste and metals, and the operational
conditions. Results suggest that combustion temperatures and retention times are the dominant parameters determining the volatility of metals in the first combustion chamber. Targeted metals were
generally partitioned in the rank of bottom ash, filter cake, fly ash and flue gas. High filter cake/fly
ash ratios showed that high temperatures in the second stage increase both the formation of gaseous
metallic compounds and the enrichment of metals in fine particles. Since ESP could not be effective
in removing fine particles and volatilized metallic compounds, the necessity of an additional system
that would remove heavy metals efficiently was emphasized for the modern incinerators.
2003 Elsevier Science B.V. All rights reserved.
Keywords: Incineration; Hazardous wastes; Heavy metals; ESP; Wet scrubbers

1. Introduction
Although the incineration technology has been used for the disposal of municipal and
hazardous (industrial or clinical) wastes for a long time in the world, it is rather a new topic
for Turkey. The first, and for the present, the only incinerator in Turkey is Izmit hazardous
and clinical waste incinerator (IZAYDAS) which was built in Izmit, the most industrialized
area of Turkey. IZAYDAS incinerator was built by Lurgi GmbH, Germany in Solaklar
Village, Izmit and started to operate in December 1997. At this time, some strict legislative
controls had been introduced by the European Commission due to the environmental risks

Corresponding author.

0304-3894/03/$ see front matter 2003 Elsevier Science B.V. All rights reserved.
doi:10.1016/S0304-3894(03)00009-8

90

M. Bakoglu et al. / Journal of Hazardous Materials B99 (2003) 89105

caused by the incinerators and the public concern over emissions of pollutants such as
particles, heavy metals, acid gases, and especially dioxins. These strict legislative controls
and emission limits have been adopted by Turkish Ministry of Environment.
This study aims to evaluate the partitioning of eight targeted heavy metals (Cr, Mn, Cu,
Pb, Sn, Co, Ni and Zn) into flue gas, filter cake (FC), fly ash (FA) and bottom ash (BA)
by presenting the results obtained during five trial burns conducted at different dates in
2000. These tests have been carried out in the plant since it started operating, as a legal
requirement for licensing by the ministry. The main objective was to determine and assess
the emissions (especially PCDD/F and heavy metal releases to atmosphere) of the plant,
but detailed analyses of wastes and solid residues were also performed to observe the fate
of metals, halogens, etc. in the incinerator.
Since the partitioning of heavy metals in the incinerator systems depends on the combustion conditions like temperature, retention time, etc. as well as some physico-chemical
properties of heavy metals, similarities and differences in the behavior of heavy metals in
pilot and real incinerator studies have a great importance both in predicting the behavior of
metals and in designing incinerator systems.

2. Information about the plant

Schematic diagram of the plant is shown in Fig. 1. The capacity of the plant is 35,000 tonnes
per year. It has a two-stage combustion system consisting of a rotary kiln and a vertical shaft.
Industrial and clinical wastes in solid and liquid forms are combusted in the plant. The dimensions of the rotary kiln are 12 m in length and 4.2 m in diameter, with combustion
temperatures in the range of 9501050 C (set point is 1000 C) and a retention time of
22.5 h for solids. The slag produced in rotary kiln is taken to an ash quench chamber to
contact with water for 10 min, and then it is transferred to landfill site by a bunker after
sampling. The second stage combustion occurs in the vertical shaft with 12 m height and
4.1 m diameter. This unit serves as a post-combustion chamber, equipped with independent
burners and fans, ensuring complete destruction of hazardous organic compounds. The
gases produced in rotary kiln, vapors of ash quench chamber and some liquid wastes are incinerated in this vertical shaft, with a retention time of about 2 s at 11501250 C (set point
is 1100 C). Air pollution control system consists of an electrostatic precipitator (ESP) and
two-stage venturi scrubber. An additional activated carbon unit for the removal of organic
gases, especially dioxins and furans was built, but it was not in the operation during the tests.
Stack gases leaving the ESP enter the first stage venturi scrubber. In the scrubber, HCl, HF
and heavy metals together with the fine particulate matter escaping ESP are removed in the
acidic environment (pH range is between 1 and 2). A two-stage pulverization washing is used
here and the utilization rate of the washwater is 170 m3 /h. Then the flue gases enter the second
stage wet scrubber in which the gas comes into contact with the lime solution again to remove
SO2 , organic gases and any remaining pollutants. Flue gases pass through two washing steps
in two different levels and 500 m3 /h washwater is used in each step. The wastewaters from
the scrubbers are treated chemically by coagulationflocculationsedimentation processes.
The solid residuals generated by the incinerator are classified in three parts: (i) bottom ash
representing the mixture of slag of rotary kiln and bottom ashes of vertical furnace; (ii) fly

M. Bakoglu et al. / Journal of Hazardous Materials B99 (2003) 89105

Fig. 1. IZAYDAS incinerator process flow scheme.

91

92

M. Bakoglu et al. / Journal of Hazardous Materials B99 (2003) 89105

ash containing the particles collected from electrofiltration of flue gas and the particles accumulated in boiler; and (iii) filter cakes, which are produced by the chemical treatment of the
wastewaters from both scrubbers. Since the incineration of most wastes takes place in rotary
kiln, almost all of the bottom ash is comprised of the slag of rotary kiln. Metals removed in
two-stage scrubbers are both in gas form (as vaporized organometallic compounds, metal
chlorides, etc.) and in solid form, as bound on ultrafine particles escaping ESP.
These solid residues are, in general, heterogeneous materials resulting from complex
processes occurring during the incineration and the raw gas treatment. ESP ashes contain
original fuel materials that have been mechanically transferred into the flue gas as well
as condensate species found on the surfaces of fly ash particles, which result from the
condensation of volatile species during the cooling phase of the flue gas inside the boiler.
Filter cakes or wet scrubber residues contain salts from the neutralization of acid gases,
mercury-bearing compounds, and other volatile-rich metal compounds depending on the
particle slip of the filter. Since fine particles that mainly pass the filter are enriched in volatile
metals, filter cakes determine the efficiency of the process, and are therefore important to
characterize.

3. Experimental methods
Metals were analyzed in bottom ash, fly ash, filter cake and flue gas. Analyses in bottom
ash, fly ash and filter cake were made by inductively coupled plasma atomic emission
spectroscopy (ICP/AES) (Spectra) at IZAYDAS Laboratory after acidic digestion based on
the analytical method of DIN 38406-E22 and expressed on weight basis (weight of metal
to total dry weight of sample). Samples of bottom ash, fly ash and filter cake were taken
according to TS 12090 and ICS 13.03.10. For each group, a sample of 2 kg was taken,
dried for 24 h at 105 C and homogenized by mixing. Then 3 g samples of each group were
digested with 50 ml HNO3 :HCl mixture at 3:1 by volume for 24 h before ICP/AES analysis.
Flue gas was sampled according to US EPA MM5 method. Sampling train included a
stainless steel probe, three impinger solutions of 100 ml 0.1 M HNO3 (submerged in an ice
bath), silica gel and an isokinetic sampler (ZAMBELLI 6000 PLUS). A flue gas volume
of 2 Nm3 was taken in 34 h in each test. Samples were analyzed directly by ICP/AES and
results were expressed as mg/Nm3 (11% O2 ).
Analysis of waste composition was made in IZAYDAS Laboratory. Samples from solid
wastes were taken according to TS 12090 and ICS 13.03.10, while samples of liquid wastes
(special liquid wastes, waste oils, etc.) were taken according to TS-ISO 5667-510. Metals
were analyzed in the same manner for each waste type. Chloride, sulfur, fluoride, iodide,
bromide and phosphorus were analyzed by ion chromatography based on DIN EN ISO
10304-2 method.

4. Results and discussion

Partitioning behavior of metals mainly depends on combustion temperatures, but S/Cl
content of wastes, other operational conditions and physico-chemical properties of heavy

M. Bakoglu et al. / Journal of Hazardous Materials B99 (2003) 89105

93

metals and compounds formed during combustion could have a significant role. Therefore,
the results of the performance tests should be evaluated separately by taking into account
the characteristics of wastes and operational conditions in each test with respect to reaching
general results.
4.1. Waste composition
A general summary of waste properties, amounts, combustion temperatures and flue gas
compositions in tests are shown in Table 1. Industrial wastes are various solid wastes coming
from different industries and fed to rotary kiln by bunker. Their calorific values lie between
12,000 and 17,000 kJ/kg. Clinical wastes include wastes in tablet or powdered forms and
their plastic packages, with calorific values between 11,000 and 16,000 kJ/kg. Other wastes
represent certain industrial solid and liquid wastes fed in barrels. Their calorific values
change according to type of waste, from about 1000 kJ/kg for emulsion oil in Test 4 to
about 35,000 kJ/kg for waste thinner in Test 1. Waste oil fed to vertical furnace serves as
auxiliary fuel, according to its calorific value that could be sometimes as high as 40,000 kJ/kg
based on the quality and water content of the oil.
4.2. Partitioning of metals
Partitioning of metals was examined by measuring the metal contents in bottom ash, fly
ash, filter cake and flue gas. Fate and behavior of metals during combustion processes were
examined in many research studies [120]. Additionally, a number of simulation studies
have been performed by using the thermodynamic models based on the minimization of
the Gibbs free energy for a system [7,10,14,2023]. These models enable the prediction
of potential metal reactions and the existing phases as a function of process variables such
as temperature, contents of chlorine and other reactants in the system. Wichmann et al.
[24] presents a detailed review of many metal-partitioning investigations based on model
simulations and pilot-scale and real incinerator studies. The studies showed that combustion
temperature is the major parameter determining the metal volatility, although there might
be a number of other factors having significant effects in waste incinerators. Therefore,
incineration tests were classified based on the main factors affecting metal partitioning (see
Table 2).
The classification in Table 2 was made according to average values for comparison of
metal behaviors in the tests. Retention time was accepted as a function of waste amount fed
to the rotary kiln. Table 3 shows the results for metals in flue gas, bottom ash, filter cake and
fly ash. Moisture contents of the residuals ranged between 30.5 and 35.0% for filter cake,
1.5 and 8.9% for fly ash, and 28.0 and 37.7% for bottom ash. Mass balance calculations for
metals (input being the wastes and output being the sum of flue gas, filter cake, fly ash and
bottom ash) showed a good and acceptable confirmation for the reliability of the results.
Output/input ratio was between 0.73 and 1.29 in Test 1; 0.72 and 1.29 in Test 2; 1.04 and
1.13 in Test 3, 1.02 and 1.27 in Test 4 and 0.95 and 1.01 in Test 5. Metal concentrations
in the flue gas were ranged between 0 and 0.04 mg/Nm3 , which were too below the metal
emission limit of 0.5 mg/Nm3 in Turkey. In partitioning of the metals, the portion emitted
to the atmosphere by flue gas did not exceed 0.4% for all metals studied in all tests. The

94

M. Bakoglu et al. / Journal of Hazardous Materials B99 (2003) 89105

Table 1
Waste properties, operational conditions and flue gas compositions in the tests
Test 1

Test 2

Test 3

Test 4

Test 5

Average

Solids to rotary kiln (kg per day)

Industrial wastes
30910
Clinical wastes
5774
Other

39634
163
3666

13527
621
1547

20621

2268

43630
5807
2998

Liquids to rotary kiln (kg per day)

Industrial liquids
1750

1826

2957

1048

4255

Liquids to vertical furnace (kg per day)

Waste oil
10563
20662

7732

7000

11005

65951

26384

30937

67695

47993

923
959

916
1073

969
1041

1109
1116

981
1052

Waste types

Total

48997

Average combustion temperatures ( C)

Rotary kiln
990
Vertical furnace
1069
Flue gas compositiona
Flow rate (Nm3 /h)
O2 (%)
CO2 (%)
CO (mg/Nm3 )
SO2 (mg/Nm3 )
NO2 (mg/Nm3 )
Particles (mg/Nm3 )
Moisture (%)

20350
8.05
6.97
0.79
0.33
123.1
1.05
18.9

24500
7.16
6.85
0.77
0.91
111.7
1.27
17.8

10100
7.72
7.13
0.81
1.40
135.4
1.12
18.3

Chemical composition of wastes (%, w/w)b

Moisture
32.7
34.9
Cl
0.18
1.58
S
0.11
0.49
Cr
0.011
0.010
Mn
0.032
0.050
Cu
0.212
0.175
Ni
0.114
0.125
Pb
0.023
0.033
Sn
0.010
0.006
Co
0.010
0.012
Zn
0.820
0.720
a
b

36.1
0.31
0.81
0.024
0.062
0.320
0.181
0.095
0.015
0.028
0.621

14900
7.35
7.20
0.83
0.89
127.2
1.03
19.1
36.8
0.43
0.18
0.022
0.056
0.080
0.045
0.024
0.010
0.012
0.446

27000
9.93
7.27
0.60
1.17
146.0
0.79
17.5
31.2
3.14
0.66
0.028
0.032
0.263
0.177
0.055
0.015
0.018
0.823

19370
8.04
7.08
0.76
0.94
128.7
1.05
18.3
34.3
1.127
0.448
0.019
0.046
0.210
0.128
0.046
0.012
0.016
0.686

Average of on-line measurements during the sampling.

On dry basis.

Table 2
Classification of incineration tests
Parameter

Test 1

Test 2

Test 3

Test 4

Test 5

First stage temperature

Second stage temperature
Retention time
Cl content
S content

Moderate
Moderate
Moderate
Low
Low

Low
Low
Short
High
Moderate

Low
Moderate
Long
Low
High

Moderate
Moderate
Long
Low
Low

High
High
Short
High
High

M. Bakoglu et al. / Journal of Hazardous Materials B99 (2003) 89105

95

Table 3
Metal contents of filter cake, fly ash and bottom ash in tests
Test 1

Test 2

Test 3

Test 4

Test 5

Average

8820
1126
15990

4987
1517
28016

8800
606
5998

9460
711
7535

9440
1556
14313

8301
1103
14370

FC
FA
BA

0.051
0.023
0.003

0.025
0.024
0.017

0.039
0.042
0.045

0.042
0.048
0.027

0.062
0.020
0.087

0.044
0.031
0.036

Mn
FC
FA
BA

0.080
0.029
0.050

0.099
0.036
0.066

0.103
0.034
0.112

0.083
0.024
0.119

0.145
0.026
0.053

0.102
0.030
0.080

Cu
FC
FA
BA

0.552
0.520
0.301

0.502
0.320
0.317

0.397
0.321
0.746

0.089
0.748
0.113

0.829
0.568
0.596

0.473
0.495
0.415

FC
FA
BA

0.239
0.165
0.200

0.310
0.178
0.225

0.196
0.117
0.460

0.065
0.321
0.066

0.603
0.124
0.417

0.283
0.181
0.274

Pb
FC
FA
BA

0.028
0.610
0.009

0.050
0.562
0.041

0.098
0.420
0.220

0.005
0.472
0.045

0.197
0.398
0.076

0.076
0.492
0.078

Sn
FC
FA
BA

0.015
0.065
0.017

0.015
0.075
0.009

0.010
0.018
0.048

0.010
0.045
0.023

0.018
0.046
0.049

0.014
0.050
0.029

Co
FC
FA
BA

0.032
0.018
0.010

0.020
0.018
0.023

0.052
0.011
0.052

0.032
0.022
0.012

0.032
0.028
0.072

0.033
0.019
0.034

Zn
FC
FA
BA

0.421
0.520
2.220

0.325
0.785
1.550

0. 530
0.385
1.847

0. 245
0.420
1.452

0. 242
0.895
3.495

0.352
0.601
2.113

Filter cake (kg per day)

Fly ash (kg per day)
Bottom ash (kg per day)
Metal concentration (%, w/w)
Cr

Ni

metal with the highest emitted fraction was Sn with an average of 0.22%. Emitted fraction
of metals was neglected in partitioning studies.
Since waste characteristics and operational conditions were different, partitioning of
metals in each test could be different. As shown in Table 2, the most similar tests with
respect to waste characteristics and operational conditions were Tests 1 and 4. Figs. 2
and 3 show the partitioning of metals in bottom ash, fly ash and filter cakes in these tests.

96

M. Bakoglu et al. / Journal of Hazardous Materials B99 (2003) 89105

Fig. 2. Distribution/partition of metals in Test 1.

Figs. 2 and 3 show that metals demonstrated similar distribution between bottom ash,
fly ash and filter cakes in similar tests, although percentages of residuals were different
for some metals (especially Cr, Pb and Ni). Accepting fly ash and filter cake fractions as
volatile, volatilization ratio of Cr and Pb was higher in Test 1, while that of Cu, Ni and Co

Fig. 3. Distribution/partition of metals in Test 4.

M. Bakoglu et al. / Journal of Hazardous Materials B99 (2003) 89105

97

was higher in Test 4. Average distributions of these eight metals between bottom ash, fly
ash and filter cakes in Tests 1 and 4 were very close: 44, 15 and 41% for Test 1 and 45, 18
and 37% for Test 4, respectively, leading to almost equal average volatilization ratios (56
and 55%). However, the waste amount in Test 1 was about 1.5 times more than the waste in
Test 4 (Table 1), which could cause less volatilization of metals due to the retention time.
This was possibly caused by the differences in waste types in tests. In Test 1 the wastes
other than bunker were clinical wastes and waste thinner with high calorific values between
10,000 and 35,000 kJ/kg, while the wastes in Test 4 were contaminated iron sheet barrels
and emulsion oil with calorific values about 1000 kJ/kg. Distribution of heavy metals in
other tests was somewhat different due to the differences in temperature, retention time,
etc. whose effects are evaluated below.
4.3. Effect of kiln temperature on metal partitioning
Since the extent of evaporation of the metals and metal compounds in the furnace is
directly related to combustion temperature, it is the key parameter for metal partitioning
in incinerators. Highly volatile metals such as mercury and cadmium start to volatilize at
the temperatures as low as 200 C. It is believed that operating temperature in an incinerator should be higher than 850 C to minimize dioxin formation [1,3,6,2527]. However
such high temperatures increase volatilization of hazardous metals and lead to adverse environmental effects. PCDD/F emission rates of IZAYDAS and their relationship with the
combustion temperatures were reported by Bakoglu et al. [28].
Having similar properties, Tests 2 and 5 could be used to compare temperature effects on
partitioning (see Table 2). Metal distribution in bottom ash, fly ash and filter cakes in these
tests is shown in Figs. 4 and 5.

Fig. 4. Distribution/partition of metals in Test 2.

98

M. Bakoglu et al. / Journal of Hazardous Materials B99 (2003) 89105

Fig. 5. Distribution/partition of metals in Test 5.

Results indicate that lower kiln temperatures in Test 2 caused all the metals except Pb, the
most volatile one, to be retained mostly in bottom ash, leading to an average volatilization
ratio of 30%, the least value observed in the tests. Bottom ash fraction of Pb in Test 2 was
42%, and it represents the highest bottom ash percentage for Pb. Similarity of partition
behavior of all metals other than Pb in Test 2 confirms the conclusion of Corella and Toledo
[4], stating the predominance of operational conditions on the fate or partitioning of the
metals for a given test. In Test 5, 41% of the metals was transferred into combustion flue gas.
Since Cl content was high in both of the tests, increased volatilities of metals (especially
Pb, Cu and Zn) at elevated temperatures through the low-boiling point chlorides could be
expected. Temperature effect is obvious for Pb, Cu, Mn and Ni with volatilities 1.52 times
higher than those of Test 2, while other metals except Sn showed no significant changes in
partitioning behavior.
4.4. Effect of retention time on metal partitioning
Another operational parameter affecting the fate of metals is retention time of the wastes
in combustion zone. By taking it as a function of waste quantity fed to the rotary kin,
two low-temperature tests with different waste loadings, Tests 2 and 3, could be compared
to determine retention time effects. Another reason for choosing these tests in spite of
differences in waste Cl content was the similarities in types and heating properties of the
wastes used in the tests. Rotary kiln temperatures were almost the same, while the waste
feed in Test 2 was as high as 2.5 times the waste feed in Test 3. Metal partitioning in Test 3
is shown in Fig. 6.

M. Bakoglu et al. / Journal of Hazardous Materials B99 (2003) 89105

99

Fig. 6. Distribution/partition of metals in Test 3.

A comparison between Figs. 4 and 6 gives the effect of retention time on partitioning. All
the metals other than Pb and Sn showed higher volatilities in Test 3, leading to an average
volatilization ratio of 45%, as compared to 30% in Test 2. Another comparison could be
made between Test 4 with moderate kiln temperature and long residence time and Test 6
with high temperature and short residence time. Although kiln temperature and waste Cl
content was higher in Test 6, 41% of metals was transferred into flue gas, while this ratio
was 55% in Test 4. Therefore, it could be said that the higher the amount of waste, the larger
the fraction of metal that is retained in bottom ash.
4.5. Effect of waste Cl and S content on metal partitioning
Heavy metals react with other elements such as O2 , Cl and S to form different compounds
(oxides, chlorides, sulfates) at high temperatures during incineration process. Phases and
possible secondary reactions of these compounds could have a significant role on metal partitioning. The presence of waste-derived Cl enhances volatilization by formation of metal
chlorides that are typically more volatile than metal oxides, and therefore increases toxic
metallic emissions. Chloride formation in an incineration system depends on many factors:
Cl/metal ratio, temperature, affinity of metal to Cl, presence of hydrogen, carbon, sodium,
etc. [5,6,10,14,16,18,22,24]. Sulfur, on the other hand, reduces the volatilization by formation of stable metal sulfates that displace the chlorides and oxides at low temperatures; while
volatile sulfides may appear at high temperatures, particularly under reducing conditions
[6,10,18,22,24].

100

M. Bakoglu et al. / Journal of Hazardous Materials B99 (2003) 89105

In IZAYDAS tests, effects of Cl and S were not apparent due to the predominance
of physical factors including temperature and retention time. Two high-Cl tests (Tests 2
and 5) showed lower metal volatilization rates as compared to other low-Cl tests, as a
result of higher feed rates. Test 4 had the same average metal volatility rate with Test 1
with similar temperature but higher waste feed; although Cl and S contents of the waste
feed in Test 4 were 2.5 and 1.5 times higher than those in Test 1, respectively. These
results suggest that kiln temperature and retention time prevail over waste Cl/S content
in determining metal volatilization in IZAYDAS. Moreover, volatilization rates of metals are in contradiction with thermo-chemical estimates of simulation models. In Test 6,
for example, elevated metal volatilities would be expected due to the highest Cl/metal
ratios (on molar basis) and kiln temperatures in the tests; however, they showed lower
values than those in low-Cl tests. The main explanation of this discrepancy is that the
pre-conditions of these equilibrium models (complete mixing of reactants, unlimited residence time, etc.) are not reached in reality [24]. Reactor configuration is another reason, since rotary kiln is a highly heterogeneous environment causing incomplete mixing
and local differences in temperature and reagent concentrations [10,18]. Difficulties related with the incineration of solid wastes (poor mixing, requirement of long residence
times for complete combustion, etc.) should also be considered. Since solid wastes contain
most of the metals, penetration of the outside temperature into the furthest point inside
the solid, i.e. the center, could take longer times than the theoretical ones, leading metals to be retained in bottom ash, even if the physical and chemical conditions are suitable
for volatilization. Therefore, the part of metals retained in bottom ash is usually larger
than that estimated by models or found in pilot-scale tests with artificial wastes. The presence of other elements with affinity to chlorine (H, Na, Mg, K, Ca, Ba, Al, Si, Fe, etc.)
has an effect also, by decreasing available Cl for formation of volatile metal chlorides
[5,10,12,14,16,20,24]. A study on Cl-distribution in residues of IZAYDAS [29] shows that
6595% of the waste Cl is retained in bottom ash, indicating much less free Cl in the
system.
4.6. Average distribution of metals in residues
Fig. 7 shows the average metal distribution of targeted heavy metals in the tests. As
shown in Fig. 7, metals are retained mostly in bottom ash, with the exception of Cr and
Pb. Larger bottom ash fractions as compared to fractions of other residues have commonly
been observed for majority of the metals except for some volatile metals such as mercury,
cadmium, and to some extent, lead and zinc, particularly in solid waste incineration studies
[1,5,6,12,14,20].
The distribution of metals in APCE residues is directly related to the particulate removal by APCE, since most metals eventually end up in the particulate phase by condensation/nucleation processes during the flue gas cooling prior to APCE [1,3,4,18,20,21].
Previous studies showed that metals tend to be concentrated in finer fractions [12,21,24,
30,31]. Wichmann et al. [24] states that both heterogeneous condensation and reaction processes of metal species taking place on the surface of particle result in an enrichment of
metals in the fine particles, while homogeneous condensation (nucleation) enables the formation of new particles <1 m from the gas phase. Electrofiltration removes larger fly ash

M. Bakoglu et al. / Journal of Hazardous Materials B99 (2003) 89105

101

Fig. 7. Average distribution/partition of metals in tests.

particles but is less efficient for vapor and finer particles. Additionally, it is likely that some
of the fine and ultrafine particles escaped ESP remained aerosolized in the flue gas. Since
fine particles that mainly pass the filter are enriched in volatile metals and removed by wet
scrubbers, filter cakes determine the efficiency of process and are therefore important to
characterize.
All metals other than Pb showed much less ESP ash fractions (ranging between 2.4 and
16.2%) as compared to filter cake fractions (see Fig. 7). This suggests that metals at the
exit of second stage vertical shaft are generally in gaseous forms or bounded to the finer
particles that could not be removed by ESP, confirming the enrichment of metals in fine
and ultrafine particulate fractions. Such high enrichment in fine particles could be related
with two-stage combustion system of the plant. Niessen [2] reported that staged combustion
systems exhibit the fine particle enrichment process. The particles dragged by gas flow are
exposed to a second combustion process and, since the temperatures are higher in second
stage, further vaporization of the metals by the effect of temperature (decreasing the adsorption of heavy metals on particles [16]), and breaking of large particles to finer ones could be
expected. Therefore, as the temperature in the second stage increases, vaporization of metal
species (on the particles coming from first stage) and subsequently, metal enrichment in
ultrafine particles by condensation processes during gas cooling increases too. For example,
the tests with higher second stage combustion temperatures (Tests 3 and 5) produced higher
filter cake/fly ash ratios on average for all metals. Consequently, we suggest that the second
combustion stage with temperatures above 1000 C together with the metal enrichment processes is the reason of high metal fractions in filter cake representing metal removal by wet
scrubbers. On the other hand, some metals (Pb, Cu, Ni and Cr) had the highest filter cake/fly

102

M. Bakoglu et al. / Journal of Hazardous Materials B99 (2003) 89105

ash ratios in high-Cl Test 5. Pb and Cu have been reported to show high affinity to Cl, while
this is not valid for Ni and Cr [5,22,24]. Although there are some model predictions showing
significant Ni volatilization at high-Cl input at temperatures about 1000 C (in the form of
gaseous NiCl and NiCl2 ), experimental studies did not comply with these predictions [10].
Similarly, theoretical equilibrium calculations of Ljung and Nordin [22] predicted that significant amount of gaseous Cr hydroxides and oxohydroxides may form in the temperature
above 500 C, which did not supported by case studies too. However, Zn is known to be
medium volatile metal and Zn volatilization increases as Cl input increases [10,24], but no
clear effect of Cl on filter cake/fly ash ratios of Zn was observed. This can be explained by
high Zn concentrations and retention of most of Cl in bottom ash [29], which leads to lower
Cl/Zn ratios and consequently minimizes the effect of Cl on Zn. Verhulst et al. [10] states
that there could be some difficulties in complete volatilization of Zn due to its presence in
larger quantities in feed waste and stable metal oxides. On molar basis, Cl/Zn ratio ranged
between 0.4 and 7.0, and considering that available Cl (that was not retained in bottom ash)
never exceeded 35% in IZAYDAS [29], these ratios should be lowered further. However,
Chen and Wey [7] stated that formation of gaseous ZnCl2 increases after Cl/Zn ratio of 5 at
7501000 C. This means that most of Zn in the incinerator is found as its elemental form
or in the forms of complex oxides, sulfates, etc. all of which are in solid or condensed forms
up to temperatures higher than 1000 C. The studies on the chemistry and mineralogy of the
bottom ash and fly ash confirmed the dominance of these compounds [11,12]. Therefore,
we suggest that high Zn concentrations in the wastes reduced Zn volatilization during incineration. It was enriched in filter cake, in accordance with the study of Forestier and Libourel
[12].
Another comparison between the fractions of fly ash, filter cake and flue gas, which was
accepted as negligible, shows that metal removal through APCE takes place in wet scrubbers
mostly, indicating the high efficiency of scrubbers in collecting submicron particles and
also for gaseous metal compounds. The effect of lime slurry should be concerned here,
because some previous studies have shown that scrubbing of the flue gas with lime slurry
after the ESP is effective in removing over 99% of most heavy metals [1]. Overall metal
removal rates over 99% in IZAYDAS agreed with these studies. On average, wet scrubbers
removed 76% of metals in the flue gas. Fig. 8 shows the percentage ratios of elements
between fly ash and filter cakes, as a reflection of the partitioning of the metals between
particulate and gaseous hosts. The ranking in Fig. 8 demonstrates the removal efficiencies
of wet scrubbers for each metal. The distribution of metals between fly ash and filter cake
differs from the results of Forestier and Libourel [12] for Pb, Co, Ni and Cr, probably
due to the higher metal enrichment in finer particulates promoted by second combustion
chamber.
In incinerators combusting hazardous wastes of various industries and clinical wastes,
high temperatures are desirable for the complete destruction of organic matter to minimize
the formation of toxic organics (dioxins, furans, etc.) during incineration, but they also
enhances the metal volatility, resulting in an increase in the release of volatile metallic
compounds, which could be also toxic, to the atmosphere. Therefore, from the public health
point of view an efficient metal removal system is indispensable in incinerators. Toxicity of
these compounds also increases the importance of further studies on the chemical behavior
of metals during incineration.

M. Bakoglu et al. / Journal of Hazardous Materials B99 (2003) 89105

103

Fig. 8. Distribution of metals between fly ash and filter cakes on average percentage basis.

5. Conclusions
1. In IZAYDAS incinerator, metals are generally partitioned with the rank of bottom ash,
filter cake and fly ash. Pb showed the highest fly ash fraction, while Cr has the highest
filter cake and Zn has the highest bottom ash fractions.
2. Filter cake fractions between 25 and 50% indicated that metals generally showed high
volatilities as compared to literature data. These were attributed to the effect of second
stage incineration and high temperatures applied in this stage.
3. Incineration temperatures and retention time are the dominant parameters in metal partitioning. In the first stage, i.e. the rotary kiln, bottom ash fractions of the metals decreases
as the temperature and retention time increase.
4. Filter cake fractions representing the volatile part of the metals are related with the
combustion temperature in the second stage, i.e. vertical furnace, and retention time.
Increasing the temperature and retention time results in an increase in partitioning to
filter cake.
5. High-Cl input did not affect the metal volatilization noticeably due to the dominant
effects of physical parameters. But it might increase the partitioning to filter cake in the
second stage combustion, particularly for the metals with high-Cl affinity like Pb and
Cu.
6. Zn showed the least volatility among the metals concerned, although it is known as a
medium-volatile metal with a significant affinity to Cl. This was explained by high Zn
content in the waste feed, which caused lower Cl/Zn ratios (between 0.4 and 7.0).
7. Low fly ash fractions together with the relatively high filter cake fractions suggested
that a higher portion of the metals in the flue gas is in the form of gaseous species or is
bound to finer particles that could not be removed by ESP. Concerning possible toxicity

104

M. Bakoglu et al. / Journal of Hazardous Materials B99 (2003) 89105

of volatile metallic compounds, an additional removal system other than that for particle
removal is necessary to control metal emissions successfully.

Acknowledgements
Literature survey was conducted at the University of Newcastle upon Tyne, in the scope
of a Turkish Scientific and Technical Research Institution (TUBITAK)NATO fellowship
program. Authors thank to TUBITAK and NATO for the financial support. Special thanks to
Prof. G.K. Anderson (Head of Environmental Engineering Department, University of Newcastle upon Tyne) for his close interest. We also thank IZAYDAS Laboratory for numerous
analyses, and Tony Flapjack for revision of the paper.

References
[1] P.T. Williams, Pollutants from incineration: an overview. Waste incineration and the environment, in: R.E.
Hester, R.M. Harrison (Eds.), Issues in Environmental Science and Technology. 2, Royal Society of Chemistry,
Thomas Graham House, Science Park, Cambridge, UK, 1994 (ISBN: 0-85404-205-9).
[2] W.R. Niessen, Combustion and Incineration Processes Applications in Environmental Engineering, second
ed., Marcel-Dekker, New York, NY, 1995 (ISBN: 0-8247-9267-X).
[3] L.A. Ruth, Energy from municipal solid waste: a comparison with coal combustion technology, Prog. Energy
Combust. Sci. 24 (1998) 545564.
[4] J. Corella, J.M. Toledo, Incineration of doped sludges in fluidized bed. Fate and partitioning of six targeted
heavy metals. I. Pilot plant used and results, J. Hazard. Mater. B80 (2000) 81105.
[5] K.S. Wang, K.Y. Chiang, C.C. Tsai, C.J. Sun, C.C. Tsai, K.L. Lin, The effects of FeCl3 on the distribution
of heavy metals Cd, Cu, Cr and Zn in a simulated multimetal incineration system, Environ. Int. 26 (2001)
257263.
[6] F.-S. Zhang, S.-i. Yamasaki, M. Nanzyo, K. Kimura, Evaluation of cadmium and other metal losses from
various municipal wastes during incineration disposal, Environ. Pollut. 115 (2001) 253260.
[7] J.-C. Chen, M.-Y. Wey, The effect of operating conditions on the capture of metals with limestone during
incineration, Environ. Int. 22 (6) (1996) 743752.
[8] F. Hasselriis, A. Licata, Analysis of heavy metal emission data from municipal waste combustion, J. Hazard.
Mater. 47 (1996) 77102.
[9] K. Nakamura, S. Kinoshita, H. Takatsuki, The origin and behavior of lead, cadmium and antimony in MSW
incinerator, Waste Manage. 16 (5/6) (1996) 509517.
[10] D. Verhulst, A. Buekens, P.J. Spencer, G. Eriksson, Thermodynamic behavior of metal chlorides and sulfates
under the conditions of incineration furnaces, Environ. Sci. Technol. 30 (1996) 5056.
[11] K.-Y. Chiang, K.-S. Wang, F.-L. Lin, W.-T. Chu, Chloride effects on the speciation and partitioning of heavy
metal during municipal solid waste incineration process, Sci. Total Environ. 203 (1997) 129140.
[12] L. Forestier, G. Libourel, Characterization of flue gas residues from municipal solid waste combustors,
Environ. Sci. Technol. 32 (1998) 22502256.
[13] C. Holbert, J.S. Lighty, Trace metals behavior during the thermal treatment of paper-mill sludge, Waste
Manage. 18 (1998) 423431.
[14] G. Trouv, A. Kauffmann, L. Delfosse, Comparative thermodynamic and experimental study of some heavy
metal behaviors during automotive shredder residues incineration, Waste Manage. 18 (1998) 301307.
[15] M.-C. Wei, M.-Y. Wey, J.-H. Hwang, J.-C. Chen, Stability of heavy metals in bottom ash and fly ash under
various incinerating conditions, J. Hazard. Mater. 57 (1998) 145154.
[16] M.-Y. Wey, L.-J. Yu, S.-I. Jou, The influence of heavy metals on the formation of organics and HCl during
incineration of PVC-containing waste, J. Hazard. Mater. 60 (1998) 259270.

M. Bakoglu et al. / Journal of Hazardous Materials B99 (2003) 89105

105

[17] K.-S. Wang, K.-Y. Chiang, S.-M. Lin, C.-C. Tsai, C.-J. Sun, Effects of chlorides on emissions of toxic
compounds in waste incineration: study on partitioning characteristics of heavy metal, Chemosphere 38
(1999) 18351849.
[18] G. Mininni, C.M. Braguglia, D. Marani, Partitioning of Cr, Cu, Pb and Zn in sewage sludge incineration by
rotary kiln and fluidized bed furnaces, Water Sci. Technol. 41 (8) (2000) 6168.
[19] M. Wobst, H. Wichmann, M. Bahadir, Distribution behavior of heavy metals investigated in a laboratory-scale
incinerator, Chemosphere 44 (2001) 981987.
[20] S.S. Thipse, E.L. Dreizin, Metal partitioning in products of incineration of municipal solid waste,
Chemosphere 46 (2002) 837849.
[21] S.K. Durlak, P. Biswas, J. Shi, Equilibrium analysis of the affect of temperature, moisture and sodium content
on heavy metal emissions from municipal solid waste incinerators, J. Hazard. Mater. 56 (1997) 120.
[22] A. Ljung, A. Nordin, Theoretical feasibility for ecological biomass ash recirculation: chemical equilibrium
behavior of nutrient elements and heavy metals during combustion, Environ. Sci. Technol. 31 (1997) 2499
2503.
[23] J.-C. Chen, M.-Y. Wey, J.-L. Su, Two-stage simulation of the major heavy metal species under various
incineration conditions, Environ. Int. 24 (4) (1998) 451466.
[24] H. Wichmann, R. Sprenger, M. Wobst, M. Bahadir, Combustion induced transport of heavy metals in the gas
phasea review, Fresenius Environ. Bull. 9 (2000) 72125.
[25] M.-B. Chang, Y.-T. Chung, Dioxin contents in fly ashes of MSW incineration in Taiwan, Chemosphere 36
(1998) 19591968.
[26] T. van Remmen, Evaluation of the available air pollution control technologies for achievement of the MACT
requirements in the newly implemented new source performance standards (NSPS) and emission guidelines
(EG) for hospital and medical/infectious waste incinerators, Waste Manage. 18 (1998) 393402.
[27] K.-J. Shin, Y.-S. Chang, Characterization of polychlorinated dibenzo-p-dioxins, dibenzofurans, biphenyls
and heavy metals in fly ash produced from Korean municipal solid waste incinerators, Chemosphere 38 (11)
(1999) 26552666.
[28] M. Bakoglu, A. Karademir, S. Ayberk, The Evaluation of PCDD/F Emissions of IZAYDAS Incinerator, in:
Proceedings of ISWA 2002, vol. 5, Istanbul, Turkey, 2002, pp. 25352542.

[29] S. Cetin, S. Veli, S. Ayberk, An investigation of halogens in Izmit

hazardous and clinical waste incinerator,
unpublished data.
[30] J.M. Chimenos, M. Segarra, M.A. Fernandez, F. Espiell, Characterization of the bottom ash in municipal
solid waste incinerator, J. Hazard. Mater. A64 (1999) 211222.
[31] M.B. Chang, C.K. Huang, H.T. Wu, J.J. Lin, S.H. Chang, Characteristics of heavy metals on particles with
different sizes from municipal solid waste incineration, J. Hazard. Mater. A79 (2000) 229239.